I think there should be a tier below Vault, where submissions are verified for two basic things: Did they complete the game without cheating, and did they beat any known times, then they would be pretty much immediately listed on the site, even without encodes available.
This would happen regardless of how sloppy the play was, let it get obsoleted by someone else.
Just saying that the formal process for accepting TASes is really really slow. If someone does a TAS that completes the basic rules (completes the game without cheating, beating known times), it should just be listed immediately, regardless of play quality or anything else.
Even the good submissions could be fast tracked to a listing, then during the time it takes to decide what to do with a submission, before it gets its moon and encodes or whatever, have the submission (which has been confirmed to beat the game without cheating) as a listing in the database.
What I really meant is that the main site should catalog unpublished submissions at the exact instant they can be confirmed to be legit playthroughs beating existing times. There could be a link to the submission, and a time for the run. They would be marked as unpublished submissions, or Gruefood, or whatever.
It turns out checking if a run completes the game without cheating takes a bunch of time and research for games that don't have a clearly defined ending point.
I wouldn't agree with the point that our judging is "really really slow", but what purpose would such a tier or listing system even have?
Warning: Might glitch to creditsI will finish this ACE soon as possible
(or will I?)
Honestly, I do not think this is a good idea. Even the vaultiest of the vault games, such as You have to burn the Rope have some merit as TASes due to their optimization. If we strip the requirement to have movies at least appear optimized, then what's even the point? The site would have to deal with a number of suboptimal throwaway runs which clog up the database and would require extra encoding and publishing time. If you just want to see a game completed, there's plenty of sources online for that.
I do not see much of the benefit either. Is the speed of the judging process really that much of a problem for most submissions? Most runs that really linger around in workbench purgatory forever have problems with syncing/encoding. For a lot of vaultable games, defining completion isn't necessarily trivial either. Hurrying along the juding process further only produces more judging errors that have to be corrected later.
I've seen it happen many times where someone submits a run that is beaten by a run that was never submitted and published. This would provide a way of establishing a time for the game.
If you could find listings for non-published runs in the main site, that would help there. Such a listing would link to the submission, and list the time. And the listing would only be created after very basic confirmation that it finishes the game without cheating. It would be listed as an unpublished submission, with no need to have a formal encode, but could become a proper TAS listing later on.
I'm not suggesting this for weird categories, just the very basic category of unambiguously completing the game.
I'm confused as to what the point of this even is? What sort of runs are beating the submitted runs? Pre-existing runs that just were never submitted to the site? That's just a lack of critical research and I don't see how this would help tbh. Runs that happened to be made after a different run was submitted? Then what would happen to the submitted run? Take it down despite being previously "listed" because a new unsubmitted run exists? There's no guarantee that the faster run would be submitted even with this change.
[16:36:31] <Mothrayas> I have to say this argument about robot drug usage is a lot more fun than whatever else we have been doing in the past two+ hours
[16:08:10] <BenLubar> a TAS is just the limit of a segmented speedrun as the segment length approaches zero
"What I really meant is that the main site should catalog unpublished submissions"
WIP threads and usermovies.
Dwedit wrote:
"where someone submits a run that is beaten by a run that was never submitted and published"
This wouldn't be solved as the problem comes from "not knowing" the existence of it. Just because there is a list with speedrun times, it doesn't means there won't be someone else posting a TAS somewhere else (our NES Duck Tales TAS with feos had this, J.Y. posted the TAS on a minor TASing site, which posted by a viewer in the submission thread).
I'm trying to think that you are interested in:
- a game's fastest completion time (found on the internet with language barriers and other difficulties)
- the time took to beat a game (same, but you are looking for a specific time range)
Since emulation is "always" changing (accuracy, syncability, possible new techniques), this should only work for platforms with no more changes or games that already known and judges able to check if TAS cheats, which of course takes time. Maybe more than watching a TAS for the normal process...
Now back to topic, my Lynx sheet, with bolded youtube links (for good game entries) are fullplays, giving a simple "time to beat" record.
While my primary goal was to show other potential game to TAS for anyone, it's slightly went into a BizHawk testing into another "I guess I will TAS it once" project bundle.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
The three-day minimum time before accepting a submission for publication was introduced so that the public would have time to raise any objections that may affect whether the run does qualify for publication or not (and also to an extent whether it should be published to moons or vault).
(The suggestion of this minimum period was put forward when some judges were accepting some submissions for publication within hours of them being submitted, which was quite controversial in at least one case, as there was possibly bias. But that's in the past now.)
Given that there exist some minimum requirements for publication (eg. no cheating, no abusing emulator bugs that do not exist in the original hardware, etc.) I don't think it would be a good idea to just automatically accept any submission that anybody sends, no matter which "tier" they would be put into. Those requirements need to be checked (and the more people check them, the better. After all, sometimes the problematic cases can be hard to spot.)
I don't think this is as bad an idea as others have suggested, especially if it's another tier. Having said that, I feel pretty conflicted about the idea. Basically, I see it as having very little upside or downside. I think it would, on some level, be nice to have a comprehensive list of the fastest completion times for every game, regardless of level of optimization. For certain games, such as Super Mario Bros. (pretending it hasn't been TASed to death), this is a kind of stupid idea. There are simple optimizations to be made and anything less than that is just pure laziness. On the other hand, I could see a place for this tier with games like Tetris, which are heavily RNG dependent and for which perfection is often the enemy of good enough.
On top of that, however, you have issues of the extra workload for judges and a rush to produce as many runs of crap games as possible. This tier would probably need to be constructed such that it is divorced from our point system and has stricter standards on what games are allowed.
I vote "meh" on the idea. I wouldn't be against it but I'm not clamoring for it either.
I'm interested in this mostly for the purposes of the TASMania project. I'm sure there will be a number of games that would be considered un-publishable under current rules. Personally I would suggest relaxing a couple of the vault rules rather than creating a whole new tier, but either way works. As an example I would point to the rule excluding sports games from the vault. NES has its fair share of (often bad) sports games that I doubt would make moons and would then be excluded from vault.
